Atrial fibrillation (AF) is a significant public health issue due to its high prevalence (> 2 million Americans) in the general population. AF is a multi-faceted, progressive disease which can increase the risk of death, stroke, and heart failure (HF). HF occurs in 30-40% of AF patients and has major implications for outcome - 25% 5-year mortality rate despite optimal medical therapy. In direct response to these sobering statistics, our proposal promotes a new paradigm for AF management - an aggressive rhythm control in patients with reversible left ventricular dysfunction (LVD) to improve LV function and survival. Our rationale for focusing on catheter ablation of AF is based on several randomized trials, which reported that it restores sinus rhythm more effectively than anti-arrhythmic drugs and with fewer side effects. For clinical decision making in AF patients with AF-induced LVD, the distinction between reversible and irreversible LVD is important to predict which patients will experience improved LV function and survival after catheter ablation. Our proposal seeks to develop a novel MRI method to distinguish between reversible and irreversible LVD in AF patients. Among various imaging biomarkers, cardiac fibrosis, as a marker of adverse structural remodeling, is particularly useful for predicting the outcome. MRI is the only non-invasive test capable of quantifying diffuse cardiac fibrosis using a cardiac T1 mapping pulse sequence. However, standard cardiac T1 mapping has two major limitations for imaging patients in AF: sensitivity to arrhythmia and long scan time. In response to these limitations, we will develop an arrhythmia-insensitive rapid (AIR) cardiac T1 mapping pulse sequence. Our central hypothesis is that the degree of LV functional recovery after restoring sinus rhythm inversely correlates with pre-existing diffuse LV fibrosis burden, which can be quantified accurately and precisely using an AIR cardiac T1 mapping pulse sequence. This hypothesis is supported by our compelling preliminary data from an established canine model with chronic AF which show that AF induces HF and diffuse LV fibrosis despite pharmacologic rate control; and compelling evidence from a previous clinical study which reported that AF ablation improves LV ejection fraction in HF patients under pharmacologic rate control. The specific objectives of our proposal are to develop and evaluate an AIR cardiac T1 mapping pulse sequence (aim 1), evaluate the correlation between MRI and histological measurements of diffuse LV fibrosis burden in established canine models with AF-induced LVD (aim 2), and evaluate the value of pre-existing LV fibrosis for predicting the LV functional recovery following catheter ablation in patients (aim 3). The long-term goal of our proposal is to create a risk stratification system for predicting LV functional improvement after catheter ablation of AF in patients with asymptomatic LVD. By using a risk-stratified approach to AF ablation, our proposal promotes a new paradigm for AF management: an aggressive rhythm control in patients with reversible LVD, which will likely have the highest chance for success and improve survival in this patient population.